Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!tut.cis.ohio-state.edu!ucbvax!bloom-beacon!eru!hagbard!sunic!dkuug!diku!terra
From: terra@diku.dk (Morten Welinder)
Newsgroups: comp.lang.pascal
Subject: Re: TP6 - Standard Out got closed
Message-ID: <1991May14.080621.29798@odin.diku.dk>
Date: 14 May 91 08:06:21 GMT
References: <382@galileo.rtn.ca.boeing.com>
Sender: terra@rimfaxe.diku.dk
Organization: Department of Computer Science, U of Copenhagen
Lines: 26

rfh3273@galileo.rtn.ca.boeing.com (Dick Harrigill) writes:

>I've just experienced a strange error with TP6.  I'm using a function
>that I've successfully used for years to check the existance of a file:

> FUNCTION this_file_exists (VAR s:string) : boolean;
> VAR f:file;
> BEGIN
>  {$I-}
>  assign(f,s); reset(f);
>  this_file_exists:=(IOError=0);
>  close(f);
>  {$I+}
> END;

>This week, for the first time, passing a bad file name to this file
>somehow closed standard output when the close(f) statement was run.
>A successive statement such as  writeln('Hi')  yeilds a runtime error
>103 - file not open.   I have gotten around the problem by placing an
>IF test in the function and not doing the close if the file was not
>properly opened.  I am, however, intrigued by this phenomenon.  The
>close(f) function should simply (with $I-) give an IOErorr, not
>screw other things up.   Can anyone give some insight?

a) close() is an error if the file is not open.
b) try using GETFATTR to check if the file exists.